Podcast: Play in new window
Subscribe: Apple Podcasts | Spotify | Amazon Music | Android | Pandora | Podchaser | TuneIn | Deezer | RSS
In this episode of the GovFuture Podcast we feature a panel discussion from the June 15, 2023 GovFuture Forum DC event that took place in Washington, DC. The hot topic panel was focused on “Advancing IT Modernization in Government”. The panel discussed how federal agencies can and should harness the power of data to drive informed decision-making and improve citizen services. Also how federal agencies can foster a culture of data-driven innovation and collaboration as well as effectively leverage emerging technologies such as AI and automation to enhance IT modernization efforts. We also discuss the idea of risk, especially as it related to adoption of new technology. There were so many incredible nuggets of knowledge shared throughout this entire panel it’s worth listening to the whole thing. The panelists at this event represented the GAO, IRS, Dept. of Interior, and the US Air Force.
Show Notes:
Trimmed Episode Transcript: (note there may be transcription errors or mis-attributions, so please consult the audio file for any potential errors)
[Kathleen Walch] Welcome everybody. I’m Kathleen Walch, I’m an executive director here at GovFuture.
We’re so excited to have this panel today on IT modernization. So I will start by letting each of you introduce yourselves for about one minute. Share your name, agency, maybe a fun fact.
[Taka Ariga] I’m Taka Ariga, I am from the government accountability office (GAO), not as an auditor, but as chief data scientist, also in the director of our innovation lab. Great.
[Mitch Winans] Good morning, everybody. Again, my name is Mitch Winans. I’m a senior advisor with the IRS’s enterprise digitalization office. I’m a good fortune of giving a demo for one of our mobile app prototypes we have right now.
Let me see. Fun fact about me, originally from Southern California in the LA area, but been in DC for 15 years. And last week was the first time that I remember seeing a California type wildfire, smoky haze in the district of Columbia. So it was kind of a weird surreal experience. But anyways, glad to be here. Look forward to the conversation.
[Andrea Brandon] Thank you. Hi, everybody. So I’m Andrea Brandon. I’m with the department of the interior and I’m the deputy assistant secretary for budget finance grants, and positions property, small business, suspension department.
And I also have an integration and innovation office, which I call the, we’re renaming it to the BI squared. But it’s really good to be with you guys here. I am a huge trekkie, a big giant nerd.
And I love to talk about this stuff constantly. Yeah. Wow. Okay. Taka’s got me beat on that. He’s got a pin, that’s messed up. You know, really happy to be here.
[Stuart Wagner] Hi, I’m Stuart. I presented earlier, chief digital transformation officer at the department of the Air Force.
Fun fact about me. I have a three year old daughter and I’ve told her she can, I’ll basically buy her a computer which she can identify all the parts of the computer. And then we’ll build it together. So we’ve been walking around Micro Center.
She’s pointing at stuff like motherboard, graphics card and things like that. She’s really been rewarding and fun. So she’s basically there.
So we’re gonna have to, we’re gonna have to buy her all the components of the computer, build it with her this summer. All right, that’s awesome. You stood up. Hey. That’s a great coffee.
[Kathleen Walch] All right, so, Taka, we’ll start with you. In the context of IT modernization, how can federal agencies harness the power of data to drive informed decision making and improve citizen services? I’ll answer that question maybe from a GAO perspective.
[Taka Ariga] Within GAO, we have this three pronged strategy. Using data is nothing new, but in the sort of the data and age of AI, we wanna make sure that we do so at a greater speed, greater effectiveness, greater accuracy.
So within GAO, the three prong strategy involves data science, data governance and data literacy. Spoiler alert, data science is the easiest part. When you have good data, everything is essentially possible. So yes, we’re working on the technical aspects of enhancing data science capacity, but part of the thing that we’re really trying to address are the foundational issues around data governance, otherwise garbage and garbage out. How do we make data assets visible? How do we make sure that we have policy surrounding access of that information? How do we make metadata exposed in a way that user can’t trust the information? We’re even even exploring an idea, giving a Yelp review for certain data assets so that people can understand the difference between let’s say a one star data asset versus like a four point five star, depending on the usage of that information.
But the other part I’ll emphasize is the notion of data literacy. Up till very recently, we’ve been operating in this sort of deterministic world, like self-sufficiency, everything is zero than one, yes or no. And it’s sort of the age of machine learning we’re talking about 62 % likelihood that somebody committed fraud.
62%, is that a yes, is that a no? So does your employee know how to interpret that information to try whatever that you’re doing, policy making, auto decision evaluation, et cetera. So I think digital literacy is an important part of not just using the data, but how you interpret the information coming out of that narrative.
This may be a controversial statement. I never mentioned analyzing data science as a sort of an exercise of seeking truth, right? Give me a set of data, I will spin any narrative you want me to do. The idea here is how do we then use that data to support the kind of policy direction that we want to get to the kind of evaluation sort of the quality that we want to achieve.
And so that certainly has a technical element of it, but data literacy helps us to then take action based on that. Yeah, I like that you say that. I think that those are really important. Our data footprint in general, people are understanding that now and it’s like, you can take that data and you can spin it in any direction, right? So you can, you know, interpretation of that.
[Kathleen Walch] So those are great things. Mitch, the next question is for you. How can federal agencies foster a culture of data driven innovation and collaboration to accelerate IT models and initiatives?
[Mitch Winans] Great question. Thanks, Kathleen. And Taka definitely wins the sock game today. I love it. I love it. I like it.
I like it. But yeah, thanks, Kathleen. It’s a great question. I think, you know, a few things that come to mind for us, I think that really focus on, we call questions zero. What are you trying to accomplish?
You know, what’s the point, figure out the why and the why based off the data or the information that you have, you know, is it justifying a certain direction, you know, what’s the problem you’re trying to solve and actually have the support for that so you can make a good data driven decision. And then from there, I think executive sponsorship is really, really key. The last three years that I’ve been with the digitalization team at the IRS, I’ve just seen how that holds true. I was actually reading a change management study recently and it said the number one factor for success of a project is an active and visible executive sponsor, which I thought was really interesting.
It was a good reminder. We’re very fortunate to have that with Harrison Smith and Molly Kay and our execs and our team as well as Commissioner Danny Werfel and some other folks across the IRS. Beyond that, I think really finding champions is key to creating that forward thinking culture at all levels, not just the senior level, but people actually working in the trenches every day. And I think that I think that with that, it’s, yeah, well, yeah, kind of a section of that.
But I think another area kind of from there that we think about within our team, we have some core cultural principles that again, got to give kudos to Harrison Smith, our leader for thinking of them. But the first one is to be pro digitalization, not anti status quo. So really try to acknowledge some of those challenges that are out there, but be positive about it, be approachable about it. Another thing is to be transparent and acknowledge those challenges. A third principle is to find balance. The work is really hard, but it’s worth it.
Then we got to stick with it together. The fourth one is to build partnerships, events like this, right? I mean, externally, we’re literally not mandated or she or not authorized or appropriated to do our jobs on her own. We need to partner with private sector, with academia, with other partners to be able to do our job and get it done, but also internally make sure we’re finding the right program offices and the key people that can help us build that culture and put that diverse team together to make it happen. And then the last cultural principle and highlight is to be kind. A lot of the work can be very frustrating and a lot of people have some very important perspectives and backgrounds for where they’re coming from. So just approach everything with empathy and respect. And I think that’s a key, those are some key ingredients for us to be successful with that.
[Kathleen Walch] Yeah, I love that one. We had a podcast actually with all of our panelists, they will be publishing and we talked about those in greater details. I encourage everyone to check that out if you’d like.
But I like how you also talked about top down, right? And really need exact news. You need champions. You need people saying, yes, please do this. I’m encouraging you. I’m fostering this innovation, this culture. It’s super important.
Andrea, the next question is for you Andrea. You know, one thing that we like about GovFuture and GovFuture Forum is it’s able to bring the entire public sector ecosystem together because you can learn so much from what others are doing and say, how can I adapt that to what I’m doing? So how can federal agencies effectively leverage emerging technologies such as AI and automation to enhance IT modernization efforts? And maybe how have you looked to other agencies?
[Andrea Brandon] So that’s a really great question. I’m going to kind of piggyback off what Mitch just said. So DOI, I was hired at DOI in 2019. And from there, I was a deputy assistant secretary over at HHS. And I was doing artificial intelligence over at HHS. And I was doing blockchain distributed ledger technology. And we hadn’t quite approached the RPA.
We were looking at that, you know, at HHS, et cetera, at the time. This is like 2016 to 2019. OK, so and let me tell you, back then, 2016, people were rolling their eyes at me about AI. They were like, whatever, Andrea, what are you talking about?
You know, but I just kept plugging along. So when I was hired at DOI, I was taking some of the things that I did over at HHS, lessons learned partnerships that I had made at HHS and bringing them over to DOI. Now, the really cool thing is when I got over to DOI, they were already doing RPA. So the issue that I was having at HHS, getting them to accept RPA, I didn’t have the issue really getting them to do the blockchain, distributed ledger technology for acquisitions, et cetera.
But when I got over to DOI, I was like, I want to do AI. This is let’s do this. And I was trying to give them the vision statement for that, you know, trying to give them the big vision. And they were like, we’re doing RPA. And, you know, we’re they were at the very beginning of it.
And so they were like, you’re the you’re our leader. Can we get your buy in? Just like you were saying, you know, can we get your top down buy in so that we can continue? Or are you going to push us in a different direction trying to push your AI? You know, so I’m for innovation.
RPA works for me. Let’s do it. And so they were so ecstatic. And we’ve been moving forward with the RPA at DOI.
Now. I’ve been going around talking to all of those partnerships. I sit on various government-wide committees. I am the co-chair for the Chief Acquisition Officers Council.
So I get a huge opportunity to talk to all your senior procurement executives and all of your heads of contract activities across the federal government. And we share a lot of information and different technological advancements that we’re doing across the government. And we also do demos. We do demos at the OMB level, et cetera. We go out to different agencies. They come into us and do demos.
And then we also partner with the private industry. So I get lots of demos coming in from different vendors to show us what’s out there, things that we had no idea that were existing at the time. Also, I have a, I don’t want to say which vendor, but we’re going out to my team at DOI is currently going out June 28th to one of the vendors to their Innovation Lab. So that we can see from them all the new and exciting innovations that they have. And then we have within DOI, we have a huge compendium of use cases for various, for various business processes where we’d like to see more innovation and more technological advancement, getting rid of Excel spreadsheets, getting rid of those PDFs, you know, that live on people’s personal computer, desktop drives and so forth. And when they retire, the information just disappeared with them. So we’re trying to fix that situation. You wouldn’t believe that we’re still basing that in 2023, but we are. And it’s partly, you know, trying to get everybody on board with the newer technology. So we have kind of our tentacles moving in different areas.
But I want to kind of also talk a little bit about the leader, the head, the top person coming to top down, giving a vision to the organization. So one of the things I did, I’m not going to kind of, I’m really jealous about that Star Trek fan that he has. So I’m going to be tracking something similar down. Yes.
And let me tell you why. Because one of the things I gave, one of the business statements I gave to DOI when I got there, and to my business integration office, which will be soon the business integration and innovation office, the BI squared office, that’s the office that has all of our technological support, they do all of our financial systems, our procurement systems, grant systems, that any integration, etc. The RPA, we’re going to need to look at a blockchain for supply chain management, etc. Okay, I gave them the vision, or I gave that director, and then she gave it to her crew. I gave her the vision that I am Starfleet, and she’s a starship, you know, she’s a female, so she’s a female captain, Kirk S. Okay, anyways, and that, you know, she’s the boldly go where no person has gone before. And we have two offices that she’s in charge of, one is in Colorado, and then one is in rest of Virginia. And so initially, you were planning to turn our Colorado office into the DOI Innovation Lab. But that’s kind of a ways away from headquarters, a lot of our employees and staff and functions are in Washington, DC, Virginia, metropolitan area. So instead, we’re planning to turn the rest in Virginia office into the Innovation Lab. So I, and they are marching forward, let me tell you, we’ve got some really good stuff happening, not just RPA, like really, and they’re very excited. And I’m going to go out and get them all Star Trek pins.
I take that challenge. Yes. And this is what we’re talking about, like sharing of ideas and, you know, and people actually get more motivated when they can see when like you guys did demos for us today, the three different demos, right? When they can see what we’re doing, and they can get in there and even play with it. You know, and we have a few, like I said, we’re looking at something called smart parks, you know, and that’s with a virtual reality. Let me just put that way, augmented reality with regard to our National Park Service, which is under the DOI. So we’re working on something with that pretty cool. But we want to turn the rest in office into an innovation lab, where it’s not just about AI or RPA or, you know, distributed ledger technology, but it’s also like you walk in here as an open environment, right? When you walk in, we want smart boards. When you walk in, we want, I want to be able to say, hey, Alexa, what’s the temperature? And then Alexa actually talks back, it’s hooked into the system, you know. I don’t want to hold a microphone.
I want to be able to talk. And it’s like, you know, so we’re working on all of that. It’s not just the system, the business process system that we’re working on. We’re also working on the infrastructure. So we’re looking at turning the rest in office into all of that with smart glass and all kinds of yummy things. Yeah. So, and I promise you guys, when we get it all up, we’ll invite all of you.
[Kathleen Walch] So anyway, with all that said, I’ll move up. So we do love field trips. And I think that’s super important too, because how can other agencies learn from what you’re doing, right? And you can be that prototype, say, this is what we’ve done. And look at what we’re able to do.
And you should bring this on board too. If we did it, you can do it. We love that.
We obviously love demos. So Stuart, the next question is for you. The Air Force is continuing with its IT modernization journey, but we do need to be careful when we’re striking this balance between security and innovation. So how are you approaching that Security and innovation?
[Stuart Wagner] Okay, two thoughts, two thoughts on that. First, I’m going to take a slightly different kind of view than Mitch’s team has taken. I think being empathetic and kind and working within the system is definitely a good thing, probably in the majority of cases, but in a minority of cases, sometimes you need to take a different perspective. Perhaps I can be messaged in empathetic ways, but I tend to find that my most ambitious projects actually come from a perspective of malicious compliance. And so working within the system, sometimes you have to think, that’s the board.
That’s the board. Can’t stop her about that. So I mean, that’s where better than I came about.
That’s actually where the projects kind of came from. It was like, hey, we have this policy. We’re not really using it.
Why don’t we just make everybody use it and then figure out what’s wrong with the policy and fix it? And so that comes from a perspective of malicious compliance. And so how do you balance between basically security and innovation? One is thinking about how do you comply, but also at the same time perform or achieve the mission you’re thinking to achieve?
And you can message that in a friendly way. Sometimes you need to say, we need to actually make a change. So the second way I think about the perspective of basically balancing between security and innovation is risk. And I think that we’ve mal-assessed risk. So one way to think about risk is opportunity costs. So for every action you take, you’re determining not to take an alternative action or an alternative course. And opportunity cost is basically the cost of not taking that action. And so inaction is an action, actually.
It’s the determination not to change. The way I see security in the DOD, we assess, if you look at risk breaks down, all these controls for the risk of an action, the risk of taking an action, what it doesn’t assess is the risk of not taking that action, the risk of remaining steady state. Sometimes the risk of what you’re currently doing is larger than the risk actually of a potential action. For example, we assess risk, you know, basically the risk of deploying software or the risk of deploying an LLM. That’s a big conversation right now, we can talk about LLMs. Sure, there are many risks with deploying an LLM to the enterprise, maybe which we can’t even assess.
There’s also a lot of uncertainty, different than risk, often associated risk, but there’s also a lot of uncertainty around it. We don’t understand how it will exist within the DOD. But there’s risks of not deploying it too. The risks of not deploying it are that our adversaries could deploy their own LLM. They could automate their own systems faster than we can. LLMs are able to produce code, they could produce extensive amounts of code that automate their enterprise and allow them to make decisions faster than us. The risks to that are unassessed in the risk management framework today. In determining whether or not to deploy software or to deploy hardware or to make change versus not, I think we have not yet assessed risk and that effectively poses a challenge to innovation. Those are the ways I think about it. I don’t have a solution necessarily other than to try to encourage decision makers to consider the risks of inaction and to take a perspective that encourages different thinking and change.
[Kathleen Walch] Thanks. We talked about this on our podcast with you about maybe being perceived as risk averse. You’re like, we are a defense agency. I do not know why people say we are risk averse, but I think that sometimes it can be perception. Sometimes it’s well, maybe you’re risk averse with adopting new technology because of the potential impacts. To get different perspectives, how do you see your agencies with adopting new technology with IT modernization and how are you weighing those risks versus rewards of moving forward in that direction?
[Taka Ariga] I guarantee GAO is more risk averse than maybe DoD. Trying to innovate among the auditors in some respects is better than oxymoron. Auditors are known to come after the fact. We tell you exactly what you did wrong after you implemented something.
That doesn’t work categorically in AI, blockchain, cloud services, sense. By the time we issued that report 12 months later, the rest of the world would have moved on to something else. To us, the exploration of foresight when it comes to data science and emerging technology is paramount to existential threats. If we don’t stay relevant, well, by definition, we become irrelevant.
It is very difficult organizationally for GAO to think about these because every organization have the day-to-day fires. “My Zoom is not working.” I don’t know. It is part of having that conversation. We talked about some form of top-down leadership to make sure that we are at some level of resources towards forward-looking aspects so that we can understand the consequences, we can understand the impact not only to operation, but for GAO specifically, it is a bit of a dual mission. We need to figure out how to audit RPA, AIs, but we would like to use the technology ourselves to enhance our oversight capacity and audit our skeptical punch. The way we drive innovation has to be deliberate and methodologically aligned. My approach is always be the discomfort, push the boundary, but importantly, never break the envelope because cleaning up on IL-5 is even Yeah, well said.
[Mitch Winans] Great, great thoughts. Yeah, just to build on what Taka and Stuart were saying, I think that a challenge we have not only the IRS just across the entire federal government is that comes down to incentives. I think we generally have a culture that rewards compliance and rewards risk aversion.
We do not have a culture policy environment that rewards good customer experience or employee experience or sound business judgments. And like Stuart was saying, I mean, uncertainties are different from risk. They often factor into risk and when it comes to contracts with private sector partners, if there is uncertainty, they’re going to embed that in their rates. You’re going to see that in the price and we obviously end up paying that as an agency or as taxpayers. So I mean, that’s an important factor for us to consider.
But I also think kind of like what else has been said, there are a lot of bad actors out there. They’re very savvy, they’re very fast and they’re quickly learning about things like generative AI and large language models and how to use those in negative or illegal ways. So the IRS and the entire federal government thinking from the criminal investigation side of the IRS, we have a responsibility to intelligently explore those technologies, get more understanding of them. If we don’t have the expertise in house that we need to partner with private sector organizations that do that can help us prepare for mitigate potential issues from bad actors that are using these tools. So I think that’s an important part.
We think a lot of the IRS with identity theft tax fraud tax evasion, racketeering some other organized crime. But then there’s also just a lot of other potential good positive use cases and we might not know how to assess that yet because we haven’t tried it. But I mean, there are there are a lot of other, you know, countries, national governments, I was just reading about Japan, the Tokyo municipal government is trying to use a they’re going to use chat GPT for informational FAQs on some of their websites, they’re they’re trying it out in a relatively safer environment, a low risk environment, it’s not risk free, but they’re taking a calculated risk to explore and try using these these new tools that are out there so that we can we can get ahead of it. So I think that’s what we’re trying to do with the IRS and what we need to do more of and then kind of looking forward to seeing how the next few months ahead go.
[Andrea Brandon] So basically within DOI, we do a lot of pilots in a sandbox, because we’re allowed to do the pilots without having them be fed ramped verse. So one of the key things as far as us then moving out of the pilot phase into production implementation is then getting it fed ramped and then that can be a little sticky. We appreciate the fed ramp process we know it’s very you know it’s it’s there for a good reason to protect us. As far as security cybersecurity, etc.
But, you know, try to move beyond the fed ramp process is something that we plan for. We were very closely with our CEO. He’s, you know, my team’s always works when I need to get in touch with him, let me tell you. But yeah, we at DOI are very collaborative with our OCR office and but we do a lot of files in the sandbox. In addition to that, we do have a systems it roadmap for all of the offices and business functions that are under me and under review but we also are building out the department’s innovation roadmap where we’re looking at the different types of innovative technologies. You know, and chat GPT where you know Microsoft just just announced their public version of chat GPT so we’re looking at that I immediately got on the teams with the CEO about that one.
Yeah. But definitely making sure that we follow security protocols is very, very important. Definitely looking at the risks.
Love what you were saying, because I have a lot under my purview. And we look at the not just the line of business for budget or finance for fans or acquisitions of property but the integrated risks. And we’re looking at also the programmatic risks and and other things and you’re right we look at the risk we don’t look at the risk of not taking an action so that’s something that we’re going to go back and look at our risk profile for you know we have some other discussions for that so that was pretty cool for this morning but yeah definitely working with our CEO office and trying to get things to be ramped is something that’s very important. Do you have any other thoughts you want to add for open it up to the audience.
I think it’s super interesting. Andrea that that you mentioned the sandbox. I think that that one of the things we’ve done at the V is from organizing hackathons which I’ve talked about previously, which basically is a sandbox with weapons system data and and folks have a lot of fun with that. And then, and then figuring out how do you then at or for fed ramp those things right so to increasingly messing with automation. Right so you can go from like product rapid prototyping to production while managing for cybersecurity risk using automation rather than human beings is how I kind of think about it and so it sounds like you guys are ahead of us that never got excited to hear about the same box. Perfect well it’s time to open up questions to the audience anybody have any questions. Good stuff here I just like I love this lab I’m actually just I’m just actually kind of curious about about this innovation lab you’re building here and of course here at rest and Virginia so I’ve been not asked more questions about, but sort of like where did that come from.
Are there. How are you interactively looks like you have an innovation lab as well and that makes it maybe talk a little about that and kind of like sort of long term goals of that I know that wasn’t necessarily not good but yeah so again can you guys still hear me okay so um I have gone out personally and looked at a lot of different innovation labs from different vendors and it’s pretty cool to actually get in I’ll tell you one of them uh we went to see this particular vendor and they were responsible for all for upgrading all of the airlines and all of the planes and so forth so in their innovation they literally had a cockpit with all the like really awesome technology and we could get in the cockpit as participating in their innovation lab and we could fly the plane now I wanted to do it because you know I’m just going to jump in and do it and your brain cannot tell the difference between augmented reality and reality I’m just going to tell you that right now this is I got in that thing my co-pilot had no idea how to fly the plane either did I so the vendor person was walking us through okay this is how you glide it on the runway you know take it off okay so that was all good we’re watching the you know the uh the the devices the the meters and the data on the dashboard and we’re looking and it has like windows like you’re looking at the plane you know and we’re looking at oh my god it’s cool and they actually mapped in real like um atmospheric things and and the the actual things like we were taken off from California so you could actually see the buildings and so forth they look like we were really taken off in the plane and we were really in the cockpit okay we’re flying we’re flying over the ocean it’s all going good a little bit tipping the plane okay no we got it straight and they’re walking us through a beginning you know and when it came time to land that thing we could not land the plane so and that
You would think oh that’s not a big deal Andrea well we couldn’t land it the plane started going sideways like we were in a storm or something I started getting sick like my I started getting like sick and I was like I started like and everybody watching us was laughing like they were like yeah but I was actually getting sick like you know because my brain couldn’t tell it was a simulation it couldn’t so that’s how real it is participating in those innovation labs it’s really cool we ended up crashing the plane we could not land it and you know luckily it was a simulation but and I couldn’t wait to get out of the seat because I wasn’t feeling well and they got me water and everything but it was really something that stuck with me like forever to be able to actually hands on do the the demo another innovation lab I went to I was actually able to put on the virtual reality and and they were filming me right I don’t want to say which finger it was but they were filming me and and they opened up the app for gaming and I don’t usually do gaming like I’m too busy doing other stuff right so okay they said oh we you do you think that was cool opening up like YouTube and watching a 3d or whatever.
We got something really great for you so we opened up a game and they were filming me right but in my when I was looking at it was real like I could actually go around the corner and open drawers and I was like really doing all this cool stuff in the game but I was fighting aliens in the game and I was screaming I was running around and you mark the room you know so I was running around the room you know because in what I could see what I was in the planet I was you know like I was in the game and I was literally fighting these aliens and when I finished they sent me the video and I looked like an idiot because nobody could see what I could see in the virtual reality you know and they were watching me they were saying that and they were laughing and everything but those things you know and other people for my team got to experience it as well and there was other things I got to experience so far I can’t we can talk all day but visiting the various innovation labs stuck with me the team members that participated that came with me it’s stuck with them and that has been a huge part of getting them to accept newer technology or create new business cases or how we can do different things with our current business processes and so forth so with that being said like I you know I said oh gosh we have an entire office that we can actually convert to an innovation lab with everything not just technology but the actual infrastructure like when you first walk in what is your experience going to be when you walk into the rest in office and we’ve done several we’ve done three visited three other innovation labs so far oh my god some really cool stuff and now we’re going to visit on June 28th another innovation lab so we’re taking that team you know every usually it’s about 10 people at a time and then we some things we film some things when I was out to film it just depends but it’s been really cool we came up with the the idea to turn it into an innovation like I said but we get it you know up and running I’ll be happy to invite one of you guys.
[Ron Schmelzer] I know that Stuart has a hard stop so just keeping an eye on that I guess I guess maybe on that question I’m going to definitely open it to the audience here I mean how in that same vein because the department of defense is on the cutting edge of the wheelio so that’s true like in terms of research and innovation our courses are like really ahead of the game because we need to be right so how does that impact And that’s kind of what you were saying earlier about risk and change. How do you bring in sort of, or do you even care? I mean, that’s been part of your line of action.
[Stuart Wagner] Yeah, I can stick around for five minutes. So one of the ways I think about, like I said, I think we confuse uncertainty with risk. And a lot of times, uncertainty is based on the inability to assess data rapidly, to understand kind of the health of our organization, the health of our systems. If we had a better understanding of the state of where we were, the decisions might make more sense, or at least the assessment of the decision might make more sense in that context and why there’s a potential, at least a request to potentially change. And so one of the things we think about at the Department of Defense, I could put a lot of something called the night after scenario.
I alluded to it earlier. The idea is basically at the start of a war, say on day one of a war with a near pure adversary, what I would suspect many commanders would seek to do at the end of that day would be to understand all of the machine and human experience that took place in that day. Everything new we saw, all the tactics that were utilized by an adversary, what are the things that worked?
What are the things that didn’t work? We would probably want to know that as soon as possible, like that day. And if you could assess and actually rapidly understand that information, that could inform, number one, where you devote your resources for additional capability development, rapid capability development, rapid innovation. And number two, what are the changes you would make to your tactics, techniques and procedures?
How would you change your strategy? And so if we’re able to make sense of our data more quickly, we can reduce uncertainty and make better decisions and perhaps be more open to small changes. And that’s agility.
What I just described is actually data-driven agility. And so anyways, that I think actually reduces our risk by actually allowing for many changes. And those changes must be based on ultimately the data undergirding our organization and the way in which we operate. So that’s how I think about changing the DOD based on data to reduce uncertainty. Hopefully that answers the question.
[Kathleen Walch] Thanks. Let’s say that I go jump in or they open it up to more questions.
[Audience Q] I have a lot of questions, cool stuff going on. So with some of the very advanced technologies that some of the different agencies are trying to implement, how are you handling finding the qualified skill within this labor market?
[Mitch Winans] All right, that’s an important and loaded question. I think that, yeah, I guess there’s a federal workforce side and then there’s the industry partnerships and contractual side. So I think on the federal workforce side, it’s tough. Kind of like I was mentioning earlier, I think federal HR has a lot of important components to it where we try to make everything transparent and fair and accessible and particularly giving job opportunities to underserved communities and particular groups like veterans and people with disabilities that’s incredibly important. There are also just some challenges with navigating the process.
It’s not very intuitive for a lot of folks. So we’ve done a lot of work at the IRS with trying to engage with college students and recent graduates and trying to think of, how can we build a future pipeline of leaders at our agency? And how do we get people interested in working for the federal government?
But maybe if they thought about different parts of the government, but they haven’t been considering the IRS or Treasury Department, kind of the financial services aspect of the government and even the law enforcement side that we have. So I think there is a public perception challenge, I’d say in general of the IRS and that we’re not just accountants and tax attorneys. We have folks like Andrew and myself that are in a completely different arena, but that message isn’t getting out there in the way that it should. So I think we need to figure out how do the people that are in particular categories of the labor market, whether they’re the data scientists or AI engineers or other parts of technology or even just communications, marketing those different areas. How do we get them interested in the IRS? How do we explain the importance of our mission?
How do we make our agency an employer of choice? Is it the training? Is it the benefits? Is it the opportunity to serve your country?
Is it something about the flexibility? Do you have autonomy with your job? Do you have other opportunities to train and get some mastery? So I think that branding communications piece is really, really important and we need to be very intentional with, if we’re trying to target a certain segment of the labor force, we need to understand how those folks communicate and receive information and become excited about things and we need to figure out how to communicate in that way with that. So yeah, great question.
[Stuart Wagner] Sure, I’m gonna share a personal story real quick to describe kind of at least anecdotal experience with the challenges with hiring. So I think we all say the right thing. Right. But I’ll share a personal story. So maybe this is now about two and a half years ago. I applied for the, so I was at the time running development and engineering for an enterprise data platform called the ban. I had about 60 engineers reporting to me for this enterprise system that ultimately is responsible for understanding, making sense of financial and enterprise data to the deputy. I was the deputy secretary of defense before that I did a software developer at Microsoft and before that I did a master’s degree in computer science, but I didn’t do an undergraduate degree in computer science. And I applied to a role that said you’re required computer science education, that specifically set an undergraduate degree in computer science, or in lieu of an undergraduate degree, you need to have kind of like the equivalent of an undergraduate degree. And I was rejected for that role, because I did not have an undergraduate degree in computer science. Even though I had been a software engineer at Microsoft, ran a bunch of developers. I think I’m somewhat technical. And so anyways, I appealed it.
Because this is where it gets really interesting, right. And they said, well, you didn’t satisfy because you need to have 30 credits in computer science related study. And I went to Penn, and Penn, it happens to use single credits. So every single credit is basically equivalent of three credits. So there’s an assumption built into the application that actually that that that every university uses the same crediting system, which it doesn’t. And so, number one, they weren’t able to assess that effectively. And then number two, they said, well, we don’t see enough calculus classes in your third credits of master’s study.
Right. I said, well, you know, but that’s that’s because, you know, advanced machine learning says CS and not math, but it’s math. Like natural language processing is is math like, like, how do you back propagation works is it’s, you know, it’s derivatives. Okay, I had to do that by hand. And so I’m fighting with a recruiter over whether or not machine learning is a form of mathematics and calculus.
Right. I ultimately had the Dean of Penn, write a letter that said, the master’s degree in computer science is more than equivalent to the bachelor study of computer science at the University of Pennsylvania. And I submitted that letter.
And they had a computer scientist, computer scientists defined as someone with an undergraduate degree in computer science, assess that letter and these statements, and they determined, I did not meet the requirements as a computer scientist. Okay. If we want to hire technical talent.
Right. This, what this means is that someone with a PhD in computer science and not study undergraduate computer science could not be considered a computer scientist according to, to opium. And this matters because even if we have existing technical talent with perhaps differing backgrounds and what is believed to be solely what is computer science, which I guess it means solely an undergraduate period computer science, then, then we can’t even hire the talent, even if they’re interested. And even if they possess the skills and experience. And so I share that because that’s what we have to break through and it’s not, it’s not about saying the right things, or even encouraging the workforce that once we have them interested, we must be able to hire them. Today we can’t. But at least my anecdote may show that we can’t. In all cases. Thanks.
[Audience Q] What’s the next step?
[Stuart Wagner] But again, like their data scientists that may have some physics, right, physics actually are very common field or math just applied math. Yeah, those are excellent that like, if the way if the language doesn’t align with like, exactly their experience like there needs to be a qualitative, there needs to be a qualitative assessment or like actually some sort of coding test to assess like we should be assessing for skills, not a degree. Thank you.
[Kathleen Walch] Right, let’s add more barriers right to get employees that we already are struggling to get so. Andrea will let you end with some parting words maybe if you have
[Andrea Brandon] I’m taking all this in because like I have a lot under my preview and we will be going back and assessing those types of things in our vacancy and then send with our new resources. So, and guess what the HR desk that the secretary says like to work out for me so I’ll be in a visiting him at the line so that was great, you know, sorry that occurred but we all learn from those past experiences.
[Mitch Winans] Yeah, just want to say thank you this was an honor to be a part of this event and this panel like so many great speakers I was also taking mental notes from Andrea, Stuart, and Take.
But this is great. Like I mentioned earlier, Andrew and I are going to stick around for a little bit if you want to see a little bit more about our mobile app prototype. And we’d love to have your feedback, play around with it, test it out. But if not, just thank you again to Kathleen and Ron and Lisa and GMU for hosting us this was great and look forward to more events like this. Yeah, this was wonderful demos were wonderful. So thank you everybody.